Client Data Hashing With Noise Injection for Private Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in ensuring data privacy and security during data transmission over networks, particularly when client devices send unencrypted data to recipient processing systems, which can be accessed by malicious actors.
Innovation Solution
Client devices perform local differential privacy operations using hash functions and noise injection to anonymize data before transmission, and recipient systems process these operations to maintain data utility while reducing communication costs and collisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If client devices transmit plain text data to recipient processing systems, then data communication is simple and fast, but data privacy and security are compromised
Solution Approach 1:
The client device performs local differential privacy operations before data transmission, including generating hash values, adding noise, and permuting values. This preliminary processing ensures that even if data is intercepted during transmission, it cannot be decrypted or exploited by malicious actors.
Solution Approach 2:
The patent introduces an encrypted output vector as an intermediary between the client device and recipient processing system. The output vector contains hashed values, noise, and permutations that serve as a secure bridge, allowing data to be transmitted without exposing original plaintext data to intermediaries or network attackers.
2Object-affected harmful factors
If client devices perform local differential privacy operations using hash functions, then data privacy and security are improved, but communication cost increases
Solution Approach 1:
The patent changes the parameter space by using hash functions that map to a reduced domain size. Instead of transmitting the full original data space, the system transmits hashed values that map to a smaller set of possible outputs, thereby reducing communication cost while maintaining privacy.
Solution Approach 2:
The system creates a copy of the data in the form of hashed values and noise-added output vectors. This copy retains the necessary information for analysis while being computationally inexpensive to transmit and process, avoiding the need to transmit the full original data.
3Quantity of substance
If hash functions are used to anonymize data, then data representation efficiency is improved, but data collisions occur
Solution Approach 1:
The patent combines multiple techniques into a composite privacy-preserving mechanism: hash functions, noise addition, and value permutation. This composite approach maintains the efficiency benefits of hashing while adding layers that reduce collision risk and maintain data reliability through the noise and permutation components.
Solution Approach 2:
The system applies different quality levels to different parts of the data representation. The hash function provides efficient compression and representation, while noise and permutation are applied selectively to maintain privacy and reduce collisions in specific data regions, allowing optimal balance between efficiency and reliability.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for hashing data on client devices. One of the methods includes generating, using a hash function, an output value that is in a hashing domain for the hash function; accessing a message size that is a smaller value than a domain size of the hashing domain; generating, using the hash function and the message size, one or more noise values; generating a message that includes the one or more noise values; and providing, to an external system, the message.